ISSN 0006-2979, Biochemistry (Moscow), 2024, Vol. 89, No. 2, pp. 279-298 © Pleiades Publishing, Ltd., 2024.
279
Enhanced ROS Production in Mitochondria
from Prematurely Aging mtDNA Mutator Mice
Irina G. Shabalina
1,a
, Daniel Edgar
1,b
, Natalia Gibanova
1,c
,
Anastasia V. Kalinovich
1,d
, Natasa Petrovic
1,e
, Mikhail Yu. Vyssokikh
1,f
,
Barbara Cannon
1,g
, and Jan Nedergaard
1,h
*
1
Department of Molecular Biosciences, The Wenner-Gren Institute, Stockholm University,
Stockholm, SE-106 91 Sweden
a
e-mail: irina.shabalina@su.se 
b
e-mail: the_daniel_edgar@hotmail.com 
c
e-mail: ngibanova@gmail.com
d
e-mail: anastasia@atrogi.com 
e
e-mail: Natasa.Petrovic@su.se 
f
e-mail: mikhail.vyssokikh@gmail.com
g
e-mail: barbara.cannon@su.se 
h
e-mail: jan.nedergaard@su.se
Received November 17, 2023
Revised January 20, 2024
Accepted January 21, 2024
Abstract— An increase in mitochondrial DNA (mtDNA) mutations and an ensuing increase in mitochondrial re-
active oxygen species (ROS) production have been suggested to be a cause of the aging process (“the mitochon-
drial hypothesis of aging”). In agreement with this, mtDNA-mutator mice accumulate a large amount of mtDNA
mutations, giving rise to defective mitochondria and an accelerated aging phenotype. However, incongruously,
the rates of ROS production in mtDNA mutator mitochondria have generally earlier been reported to be lower–
not higher– than in wildtype, thus apparently invalidating the “mitochondrial hypothesis of aging”. We have
here re-examined ROS production rates in mtDNA-mutator mice mitochondria. Using traditional conditions for
measuring ROS (succinate in the absence of rotenone), we indeed found lower ROS in the mtDNA-mutator mi-
tochondria compared to wildtype. This ROS mainly results from reverse electron flow driven by the membrane
potential, but the membrane potential reached in the isolated mtDNA-mutator mitochondria was 33mV lower
than that in wildtype mitochondria, due to the feedback inhibition of succinate oxidation by oxaloacetate, and to
a lower oxidative capacity in the mtDNA-mutator mice, explaining the lower ROS production. In contrast, in nor-
mal forward electron flow systems (pyruvate (or glutamate) + malate or palmitoyl-CoA + carnitine), mitochondrial
ROS production was higher in the mtDNA-mutator mitochondria. Particularly, even during active oxidative phos-
phorylation (as would be ongoing physiologically), higher ROS rates were seen in the mtDNA-mutator mitochon-
dria than in wildtype. Thus, when examined under physiological conditions, mitochondrial ROS production rates
are indeed increased in mtDNA-mutator mitochondria. While this does not prove the validity of the mitochondrial
hypothesis of aging, it may no longer be said to be negated in this respect. This paper is dedicated to the memory
of ProfessorVladimirP. Skulachev.
DOI: 10.1134/S0006297924020081
Keywords: mtDNA mutator mice, ROS production, aging, succinate, membrane potential, oxidative phosphorylation
Abbreviations: mtDNA,mitochondrial DNA; ROS,reactive ox-
ygen species.
* To whom correspondence should be addressed.
INTRODUCTION
Aging has been suggested to be a consequence
of accumulation of mutations of mitochondrial DNA
(reviewed in[1]). According to this, “the mitochondrial
hypothesis of aging”, these mutations would lead to er-
rors occurring in the proteins of the respiratory chain,
which in turn would lead to increases in mitochondrial
production of reactive oxygen species (ROS). The ROS
generated in this way would then lead to oxidative
damage of cellular proteins, DNA, and lipids, which
would finally manifest as aging [2, 3]. There is, howev-
er, presently no agreement concerning the validity of
this hypothesis [4-6].
SHABALINA et al.280
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
A substantial number of studies do support a role
for mtDNA mutations in aging (reviewed in [7,8]). There
are several studies showing increased ROS production
with age in both animal systems and humans [9-13].
However, not all studies agree [14, 15]). In reality, the
generally observed parallel increase of ROS and mtDNA
mutations with age has made it difficult to determine a
cause-and-effect relationship [16, 17].
In this conundrum, the construction of the mtDNA
mutator mouse [18, 19] should principally enable sep-
aration of cause and effect. Indeed, solely due to an
increased rate of mitochondrial DNA mutations, this
mouse shows features that can adequately be described
as signs of premature aging, and it also has a signifi-
cantly reduced lifespan [18, 19]. Whether this is a mod-
el that replicates the normal aging process may be dis-
cussed [1]. However, it undoubtedly opens for studies of
the processes that lead from mitochondrial DNA muta-
tions to the features of premature aging. Studies of this
mouse should then be a way of establishing the possi-
ble significance of ROS production for aging.
Thus, in the mitochondrial hypothesis of aging, a
predicted feature of increased mtDNA mutations would
be an increased level of ROS. This ROS would then
be responsible for the protein alterations that finally
manifest themselves in features of premature aging.
However, contrary to such expectations, in the mito-
chondrial mutator model, there is little evidence of
oxidative damage [20-22]. Additionally and important-
ly, when the mutator mitochondria have been ana-
lyzedexvivo, no increase in production of ROS by iso-
lated mitochondria has been reported [23, 24], severely
weakening the mitochondrial oxidative damage theory
of aging [5].
However, despite these observations, certaininvivo
investigations imply that an increased level of ROS
production is involved in the aging process observed
in the mtDNA mutator mice. For example, we have
used the mitochondria-targeted mass spectrometry
probe MitoB for estimation of hydrogen peroxide
within mitochondria of living mice. In these studies,
the in vivo level of mitochondrial hydrogen peroxide
was enhanced in old mtDNA mutator mice, suggest-
ing that the prolonged presence of mtDNA muta-
tions in vivo increases hydrogen peroxide production
that could contribute to an accelerated aging pheno-
type [25]. Additionally, when the mtDNA mutator mice
were treated with the mitochondrially targeted anti-
oxidant SkQ that would reduce ROS levels, the aging
process of the mtDNA mutator mice was markedly de-
layed [26]. Similarly, in mice overexpressing catalase
targeted to mitochondria, heart hypertrophy and fibro-
sis in the mtDNA mutator mice was attenuated [27].
These in vivo observations are thus supportive for an
increase in ROS being the cause of the aging process
and for this increase being an effect of the mitochon-
drial mutations. The reason for the apparent discrep-
ancy of the results obtainedex vivoandin vivois cur-
rently unknown.
In the present study, we have therefore re-exam-
ined ROS production in mitochondria isolated from
heart and liver of the mtDNA mutator mice. In par-
ticular, we have focused on examining the mitochon-
dria when they are metabolizing physiologically rele-
vant substrates and doing this during active oxidative
phosphorylation, i.e., we have examined conditions
that may be said to more closely reflect the situa-
tionin vivo. We conclude that under these more phys-
iologically relevant conditions, the mitochondria from
the mtDNA mutator mice do produce more ROS, and
the mitochondrial hypothesis for aging is thus not ne-
gated in this respect.
MATERIALS AND METHODS
Animals.Mice heterozygous for the mtDNA muta-
tor allele (+/PolgA
mut
) [18] were backcrossed to C57Bl/6
mice for at least 6 generations. After intercrossing
mice heterozygous for thePolgA
mut
allele and genotyp-
ing the offspring as previously described [18], mtDNA
mutatormice were identified as the homozygote trans-
genic offspring; heterozygote offspring were not used;
homozygote wildtype offspring were used as wild-
typemice. The mice were fedadlibitum(R70 Standard
Diet, Lactamin), had free access to water, and were kept
on a 12 : 12 h light : dark cycle in cages with 4-5 animals
at 24°C. The animals were routinely examined at the
age of 25 weeks, i.e., at a time point where the first
symptoms of premature aging, e.g. slight kyphosis,
alopecia, and impaired weight gain, are observable
in the mtDNA mutator mice [18, 26]. Oxidative stress
biomarkers were measured in mice at the age of 40
weeks, i.e., at a time where symptoms of aging become
grossly evident [26]. Animal protocols were in accor-
dance with guidelines for humane treatment of ani-
mals and were reviewed and approved by the Animal
Ethics Committee of the North Stockholm region.
Tissue collection and mitochondrial isolation.
Mice were anaesthetized for 1 min by a mixture of
79% CO
2
and 21% O
2
and decapitated. All experiments
were made in parallel, i.e., preparations from wildtype
and mtDNA mutator mice were directly compared on
each experimental day. Hearts from two mice were
placed into ice-cold medium containing 100 mM sucrose,
50mM KCl, 20mM K-TES, 1mM EDTA and 0.1% (w/v)
fatty-acid-free bovine serum albumin (BSA) and were
freed of white fat and connective tissue, weighed and
used for mitochondrial isolation. Liver mitochondria
were isolated as previously described [28, 29].
The tissues were finely minced with scissors and ho-
mogenized in a Potter homogenizer with a Teflon pestle.
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 281
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
During mincing and homogenizing, the heart fragments
were treated with nagarse, added to the medium at a
concentration of 1mg perg of tissue. Throughout the
isolation process, tissues were kept at 0-2°C.
Mitochondria were isolated by differential cen-
trifugation. Tissue homogenates were centrifuged at
8500gfor 10min at 2°C using a BeckmanJ2-21M cen-
trifuge. The resulting supernatant, containing floating
fat and nagarse, was discarded. The pellet was resus-
pended in ice-cold medium containing 100 mM su-
crose, 50 mM KCl, 20 mM K-TES, 1 mM EDTA and 0.2%
(w/v) fatty-acid-free BSA. The resuspended homoge-
nate was centrifuged at 800gfor 10 min, and the result-
ing supernatant was centrifuged at 8500g for 10 min.
The resulting mitochondrial pellet was resuspended
in the same buffer (but albumin-free) and centrifuged
again at 8500gfor 10min. The final mitochondrial pel-
lets were resuspended by hand homogenization in a
small glass homogenizer in the same medium. Thecon-
centration of mitochondrial protein was measured us-
ing fluorescamine [30] with BSA as a standard. Mito-
chondrial suspensions were kept on ice for no longer
than 4h during measurements.
All mitochondrial preparations were tested by
quality control traces, examining adequate respiratory
control by following the effect of ADP and FCCP addi-
tions on ComplexI substrate respiration rates.
Mitochondrial ROS net production in heart
mitochondria. Mitochondrial H
2
O
2
net production in
heart mitochondria was determined fluorometrically
by the use of theAmplex Red™reagent, Trade Mark of
Molecular Probes, USA. Oxidation of Amplex Red cou-
pled through horseradish peroxidase to reduction of
H
2
O
2
produces the red fluorescent product resorufin
[31]. Mitochondria (0.05-0.2 mg of mitochondrial pro-
tein ml
–1
) were incubated at 37°C in a buffer consisting
of 100 mM sucrose, 20 mM K
+
-Tes (pH 7.2), 50 mM KCl,
2 mM MgCl
2
, 1 mM EDTA, 4 mM KP
i
and 0.1 % fatty-
acid- free BSA. All incubations also contained 5 µM
Amplex Red, 12 units ml
–1
horseradish peroxidase and
45 units ml
–1
superoxide dismutase. The reaction was
routinely initiated by addition of mitochondria fol-
lowed by the successive addition of substrate: Com-
plex II substrate (5 mM succinate), or Complex I sub-
strates (2 mM malate + 5 mM pyruvate for heart (2 mM
malate + 5 mM glutamate for liver) or a fatty acid-
derived substrate (30 µM palmitoyl-CoA + 5 mM carni-
tine), followed by successive addition of rotenone
(1.7 µM), antimycin A (3 µg/ml) and myxothiazol (0.8µM)
or ADP (450µM).
The fluorescent signal was measured by three tech-
niques: (1)  kinetics of fluorescence emitted through
a band pass filter of 600 ± 20 nm from an excitation
wavelength of 545 nm was followed in a 3 ml cuvette
for 10-15 min (2 points persec) with a SIGMA ZWS-11
spectrofluorometer (SIGMA instrument GMBH, Berlin,
Germany); (2)  simultaneous fluorescent signals from
mitochondrial samples supported by various substrates
and ADP were detected with an EnSpire Multimode
Plate Reader (PerkinElmer, USA) in a 24-well plate. The
excitation wavelength 563 nm and the emission wave-
length 584 nm were set by applying the optimisation
mode of EnSpire plate reader. (3)  Simultaneous mea-
surements of fluorescence and oxygen consumption
rates were performed by the O2k-MultiSensor System
(Oroboros Instruments Corp.,Innsbruck, Austria).
The rate of H
2
O
2
production was calculated as the
change in fluorescence intensity during the linear in-
crease, as earlier described [32]. Calibration curves
were obtained by adding known amounts of fresh-
ly diluted H
2
O
2
(concentration of stock solution was
checked at 240 nm using a molar extinction coefficient
of 43.6) to the assay medium. The standard curve was
linear in a range up till 500nM H
2
O
2
.
Mitochondrial ROS net production in liver mito-
chondria.The measurement of H
2
O
2
production inliver
mitochondria is problematic due to the presence of
catalase in the mitochondria [33, 34]. Therefore, we
measured superoxide for examination of ROS produc-
tion in liver mitochondria. Superoxide was measured
using the superoxide-induced conversion of dihydro-
ethidium (DHE) (Molecular Probes) to ethidium at 37°C
using an excitation wavelength of 495 nm and col-
lecting the emission via a narrow band pass filter at
570 ± 5 nm with a SIGMA ZWS-11 spectrofluorometer
(SIGMA instrument GMBH, Berlin, Germany) [29, 32].
For validation of DHE as a potent tool for ROS mea-
surement, we have earlier compared the DHE and
Amplex Red methods and obtained qualitatively simi-
lar results [32].
Mitochondrial oxygen consumption. Heart mi-
tochondria (0.1-0.2 mg protein/ml) were incubated in
a medium of the same composition as that used for
ROS production. Oxygen consumption rates were moni-
tored using two techniques: an O2k-MultiSensor Sys-
tem (Oroboros Instruments Corp. Innsbruck, Austria)
and a Clark-type oxygen electrode (Yellow Springs
Instrument Co., USA) in a sealed incubation chamber
at 37°C, as earlier described [35]. The output signal
from the Clark-type electrode amplifier was electron-
ically time-differentiated and collected every 0.5 s
by a PowerLab/AD Instrument (application program
Chart v5.0.1.). Phosphorylation state 3 was measured
in the presence of 450 µM ADP. State-4 respiration was
measured as the residual respiration after addition
of 2 µg/ml oligomycin. Maximal oxygen consumption
rates (uncoupled state) were obtained by addition of
FCCP at a final concentration of 1.0-1.4µM.
Mitochondrial membrane potential. Mitochon-
drial membrane potential measurements were per-
formed with the dye safranin O [36]. The changes in
absorbance of safranin O were followed at 37°C in
SHABALINA et al.282
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
an Olis® modernized Aminco DW-2 dual-wavelength
spectrophotometer at 511-533 nm with a 3-nm slit.
Olis GlobalWorks™ software was used for recording
and quantification. Calibration curves were made for
each mitochondrial preparation in K
+
-free medium
and were obtained from traces in which the extra-
mitochondrial K
+
was altered by addition of KCl in
a 0.1-20 mM final concentration range in the pres-
ence of 3 µM valinomycin. The change in absorbance
caused by each addition was plotted against [K
+
]out
and the intramitochondrial K
+
, [K
+
]in, was estimated
by extrapolation of the line to the zero-uptake point, as
described [36]. The absorbance readings were used to
calculate the membrane potential(mV) by the Nernst
equation according to: Δψ = 61 mV · log ([K
+
]in / [K
+
]out).
Proton leak. To determine proton leak, mito-
chondrial membrane potential and oxygen consump-
tion measurements were performed in parallel using
the same media consisting of 100 mM sucrose, 20 mM
K
+
-Tes (pH 7.2), 50 mM KCl, 2 mM MgCl
2
, 1 mM EDTA,
4 mM KPi, 0.1 % fatty-acid-free BSA, 2 µg/ml oligomycin,
5 µM safranin O and 5 mM succinate, in the presence of
increasing amounts of malonate (0.5-3.3 mM), at 37°C.
An energizing preincubation time of ≈3.5 min was used
before malonate additions; proton leak kinetic mea-
surements were finalized within 11-12 min as in [35].
Protein oxidative modification biomarkers.Ox-
idative stress biomarkers were analyzed in liver total
tissue homogenate. For collection of tissue, mice were
anaesthetized for 1-2 min by a mixture of 79% CO
2
and
21% O2 and decapitated. Tissues were dissected out
and rapidly placed in liquid nitrogen; then tissues were
powdered under liquid nitrogen, weighed, divided into
small amounts and stored under nitrogen at –80°C.
A small amount was homogenized in RIPA buffer with
proteinase inhibitor (Complete Mini, Roche) and pro-
tein concentration measured.
4-HNE-adducts was detected by immunoblot analy-
sis with polyclonal antibodies from Alpha Diagnos-
tics (HNE12-S, dilution 1 : 1000) as described [37, 38].
4-HNE-BSA protein conjugate (HNE12-C, Alpha Diag-
nostics) was used as an internal control.
Carbonyl groups of proteins were analysed by
OxyBlot Protein oxidation detection kit (Chemicon
International) consisting of several steps: derivatiza-
tion of carbonyl groups to 2,4-dinitrophenylhydrazone
(DNP-hydrazone), separation of protein samples by
SDS PAGE (12%), and Western blotting using primary
antibodies specific to the DNP moiety of the proteins.
A mixture of Standard Proteins with attached DNP
residues (Chemicon International) was used as inter-
nal control.
Statistics. KaleidaGrapH 4.5.2 Synergy Software
(Reading, USA) was used for the graphs and statistical
analysis. Groups were compared with Student’s two-
tailed t-test. All data were expressed as means ± SE.
Significance was accepted at the level ofp < 0.05 (indi-
cated in the graphs by one symbol),p < 0.01 (two sym-
bols) andp < 0.001 (three symbols).
RESULTS
To clarify the relationship between mitochondrial
DNA mutations and ROS production, we examined ROS
production in mitochondria isolated from heart and
liver of the mtDNA mutator and wildtype mice. Weuti-
lized a series of protocols for examination of ROS
production and clarified the cause for the differences
seen. We finally examined whether the alterations in
ROS production were associated with differences in
degrees of oxidative damage.
Diminished ROS production in mitochondria
from mtDNA mutator mice under classical ROS as-
sessment conditions. ROS production in isolated mi-
tochondria is classically examined with the complex
II-coupled substrate succinate (reviewed in [39-41]).
Besides being a classicalin vitroapproach to ROS mea-
surement as such, succinate-supported ROS produc-
tion has attracted much attention in the longevity
field since it somewhat puzzlingly has been proposed
as both being involved in lifespan reduction [42] and
lifespan extension [43].
In isolated mitochondria, added succinate is oxi-
dized by Complex II and this reduces ubiquinone (Q)
and generates a proton-motive force that is sufficient-
ly high to drive electrons thermodynamically uphill
through Complex I, to reduce NAD
+
to NADH (reverse
electron flow). This results in superoxide production
from semi-reduced coenzyme Q (semiquinone) at the
coenzyme Q-binding site of ComplexI [44, 45]. We ini-
tially used this standard condition for ROS production
to examine mitochondria isolated from the heart and
liver of wildtype and mtDNA mutator mitochondria.
H
2
O
2
production from the heart was measured with the
Amplex Red system as exemplified with a represen-
tative trace (Fig. 1a); for liver mitochondria, the DHE
method was used (Fig. 1b) (see Materials and Methods
section).
ROS from Complex I. After addition of succinate,
wildtype mitochondria displayed a high amount of ROS
production, which was markedly inhibited in heart mi-
tochondria by rotenone (Fig.1a). A similar effect was
seen with liver mitochondria (Fig.1b). Rotenone inhib-
its Complex I at subunit ND1 [46]. The inhibition by
rotenone therefore indicated that Complex I was the
major site of ROS production in these mitochondria,
and that electrons required for ROS production are
delivered to Complex I by reverse electron transport
from ComplexII.
There was a clearly diminished rate of ROS pro-
duction in heart mitochondria from mtDNA mutator,
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 283
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
Fig. 1.ROS production in mitochondria energized by succinate and examined under traditional conditions. a,b)A representa-
tive trace of (a)Amplex Red fluorescence and (b)dihydroethidium (DHE) measurements in wildtype heart(a) and liver(b) mito-
chondria energized by succinate. Additions were 0.1mg/ml heart (HM) or liver (LM) mitochondria, 5mM succinate, 1.7µM rote-
none, 3mg/ml antimycinA and 0.8µM myxothiazol. c)Hydrogen peroxide production rate under conditions of reverse electron
flow from ComplexI (measured principally as shown in(a) and expressed as hydrogen peroxide amount, based on calibration)
in heart mitochondria isolated from wildtype (WT, white bars) and mtDNA mutator mice (Mut, black bars). The difference be-
tween the rate with succinate alone and that after the further addition of rotenone(∆) is considered to represent ROS production
from ComplexI. d)Hydrogen peroxide production rate after antimycinA addition (as in a) in heart mitochondria isolated from
wildtype and mtDNA mutator mice, the rate after further myxothiazol addition, and the ∆ between these. The rate after myxo-
thiazol represents ROS from ComplexII and the ∆ rate ROS from ComplexIII. e)Relative ROS production rates from ComplexI
and ComplexIII in mitochondria isolated from mtDNA mutator mice. ROS refers to hydrogen peroxide in heart (measured as
in a) and to the change in DHE fluorescence signal intensity in liver mitochondria (measured as in b). For each preparation day,
the∆(b) rate in wildtype mitochondria was set to as 100% and the ∆ rate in mtDNA mutator was expressed as a percentage of this.
Values in (c)-(e) represent means ±SE of 5-7 independent parallel mitochondrial preparations. *,**Indicate statistically signif-
icant differences between wildtype and mtDNA mutator mice (p<0.05,<0.01). f)A representative trace of the effect of FCCP
on Amplex red fluorescence (H
2
O
2
levels) measured in wildtype heart mitochondria energized by succinate. Additions were
0.2mg/ml mitochondria (Mit), 5mM succinate, and 1.4µM FCCP.
SHABALINA et al.284
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
compared to that in wildtype mitochondria (Fig. 1c).
The small amount of residual ROS generated in mi-
tochondria after rotenone addition was not different
between wildtype and mtDNA mutator mitochondria
(Fig.1c). We defined the Complex I-derived ROS as the
difference (delta) in ROS production before and after
addition of rotenone. ROS production from Complex I
was thus clearly reduced in the heart mitochondria
from the mtDNA mutator mice (Fig. 1c). Similarly to
the case in heart mitochondria (Fig. 1c), liver mtDNA
mutator mitochondria produced less ROS from Com-
plexI than did wildtype mitochondria (Fig.1e).
From Complex III. Another major site capable of
producing ROS under these conditions is Complex III
(forward electron flow from Complex II). To study
this site, we further added first antimycin A and then
myxothiazol to the mitochondria (Fig. 1, a, b). Antimy-
cinA is a ComplexIII Q
i
-site inhibitor that blocks elec-
tron transfer from b heme to quinone, resulting in
increased formation of semiquinone at the Q
o
-site of
Complex III, which in turn can transfer an electron to
oxygen, yielding superoxide. In heart and liver mito-
chondria, antimycin A addition accordingly increased
ROS production (Fig. 1, a, b). Also, here, ROS produc tion
was lower in mtDNA mutator mitochondria (Fig. 1d).
Myxothiazol blocks electron entry from Complex II
into the Q
o
-site of Complex III, preventing the forma-
tion of this semiquinone and hence decreasing ROS
production (Fig. 1, a, b). Thus, after myxothiazol addi-
tion, ROS production from ComplexIII is blocked, and
also reverse electron transport into ComplexI is here
blocked by rotenone (Fig. 1, a, b). We thus define Com-
plex III-derived ROS production as the difference (del-
ta) between the ROS production rate after antimycin A
and after myxothiazol (Fig. 1d). Defined in this way,
it is seen that ROS production from Complex III in
mtDNA mutator mice was 18 % lower in heart mito-
chondria. The production in liver mitochondria was
unchanged (Fig.1e).
From Complex II.After myxothiazol addition (Fig. 1,
a, b), the only residual ROS production should be from
ComplexII. This rate was equal in wildtype and mtDNA
mutator mitochondria from heart (Fig. 1d), as well as
from liver (not shown). Thus, the conclusion from
these standard experiments with succinate as electron
donor was that mtDNA mutator mitochondria showed
either an unchanged (from Complex II) or a clearly
diminished rate of ROS production (from ComplexesI
and III). It is thus obvious that the in vitro ROS mea-
surements in succinate-supported mitochondria yield
results that are in direct contrast to the results obtained
ininvivomeasurements [25].
The low membrane potential in the mtDNA
mutator mitochondria respiring on succinate may
explain the low ROS production rate.In order to ob-
tain further insight into the physiological relevance of
the use of succinate for studying ROS production in mi-
tochondria in vitro, we analyzed the bioenergetics of
mitochondria under the conditions classically used for
measuring ROS production. In particular we focused
on how ROS produced through succinate-supported
respiration is dependent on the membrane potential.
ROS production in mitochondria energized by suc-
cinate-induced reverse electron flow is susceptible to
changes in the membrane potential (reviewed in [39-41]
and seen by us in brown adipose tissue mitochondria
[32]). Indeed, the uncoupler FCCP eliminates the high
ROS production supported by reverse electron trans-
port in heart mitochondria (Fig. 1f), as expected [47, 48].
Thus, the observed low ROS production in mtDNA mu-
tator mitochondria oxidizing succinate (Fig.1) could be
due to a lower membrane potential in the mitochondria
isolated from these mice.
We assessed the membrane potential by safranin O
absorbance in mitochondria respiring on succinate
without rotenone (Fig. 2). The membrane potential was
stable for both wildtype and mtDNA mutator mito-
chondria (Fig. 2a). It was eliminated by FCCP (Fig. 2a).
After calibration of each mitochondrial preparation
versus K
+
gradients, we calculated the membrane po-
tential to be as much as 33 mV lower in mtDNA mutator
mitochondria than in wildtype mitochondria (200 mV
wildtype vs 167 mV mutant) (Fig.2b), suggesting that a
low membrane potential could be the reason for the
reduction in ROS production seen in the mtDNA mu-
tator mitochondria under these conditions: the ener-
gy may not be enough to support the reverse electron
flow to ComplexI.
To establish the cause of the lower membrane po-
tential, we first examined the respiratory capacity of
heart mitochondria under these conditions. As seen in
Fig. 2c, contrary to what is expected in isolated mito-
chondria the respiration rate of wildtype heart mito-
chondria was not limited by proton leak: addition of
the uncoupler FCCP induced just a brief increase in
respiration with subsequent inhibition, principally in
agreement with e.g., [49]. Also, in the mtDNA mutator
mitochondria, the respiration was not regulated by the
membrane potential, as FCCP addition also here led to
inhibition of respiration (Fig.2c) after a short stimula-
tion.
However, when mitochondria respire on succinate
in the absence of rotenone (as is the case in the clas-
sical design for studying ROS production used here),
there may be an inhibition of Complex II (succinate
dehydrogenase) activity due to the successive accumu-
lation of oxaloacetate in the mitochondria; this oxalo-
acetate will with high affinity compete with succinate
for the binding site in Complex II [50]. Thus, this in-
hibition may limit the rate of succinate oxidation and
the oxidative rate is then not regulated by proton leak
through the membrane.
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 285
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
Fig. 2.The effect of rotenone on succinate-supported respiration and membrane potential. a) Representative traces of mem-
brane potential measurements in heart mitochondria isolated from wildtype (thin line) and mtDNA mutator mice (thick line).
Additions were 5mM succinate, 0.2mg/ml mitochondria (Mit), and 1.4µM FCCP. b)Comparison of membrane potential levels
in heart mitochondria from wildtype and mtDNA mutator mice, examined principally as in(a). Values represent means ±SE of
6independent parallel mitochondrial preparations. *Indicates statistically significant difference between wildtype and mtDNA
mutator mice (p < 0.05). c) Respiratory traces comparing succinate-supported respiration of wildtype (thin line) and mtDNA
mutator (thick line)heart mitochondria in the absence of rotenone. Additions were as in(a). d)Respiratory traces comparing
succinate-supported respiration of wildtype (thin line) and mtDNA mutator (thick line)heart mitochondria in the presence of
rotenone (1.7µM). Additions were 0.2mg mitochondria (Mit), 5mM succinate, 2µg/ml oligomycin, and 1.4µM FCCP. e)The effect
of rotenone on succinate-supported oxygen consumption rate in heart mitochondria from wildtype and mtDNA mutator mice,
examined principally as in (c) and (d), rate after succinate addition. Values represent means ±SE of 4-6 independent parallel
mitochondrial preparations. *Indicates statistically significant difference between wildtype and mtDNA mutator mice (p<0.05)
and #indicates significant effect of rotenone (p<0.05). f)Absorbance traces comparing succinate-supported membrane poten-
tial of wildtype (thin line) and mtDNA mutator (thick line)heart mitochondria in the presence of rotenone (1.7µM). Additions
as in(a). For direct comparisons, traces shown in a, c, d and f are from one experimental day, with parallel preparations of
wildtype and mtDNA mutator mitochondria, examined in parallel for respiration and membrane potential.
SHABALINA et al.286
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
Proton leak.To establish whether the lower mem-
brane potential observed above in the mtDNA mutator
mitochondria was due to respiratory limitation, rather
than to increased proton leak (uncoupling), we made
proton leak determinations. We examined proton leak
kinetics under conditions where the mitochondria re-
spired on succinate and oxygen consumption was suc-
cessively inhibited with malonate, while assessing the
membrane potential that could be reached when the
capacity for succinate oxidation was thus successively
(further) diminished.
The separate data on oxygen consumption and
membrane potential were in themselves illustrative.
Resting oxygen consumption rates were much low-
er in the mtDNA mutator mitochondria compared to
wildtype (Fig. 3a), and the respiration in both types
of mitochondria was inhibited by the first addition of
malonate. As is evident from the data above, the mem-
brane potential was markedly lower in mtDNA muta-
tor mitochondria, but in contrast to respiration, the
first addition(s) of malonate did not affect the mem-
brane potential in a measurable way (Fig.3b).
When these data were re-graphed as proton leak
curves (“Brand plots”), i.e., as respiration as a function
of the driving force (membrane potential), remarkable
curve shapes resulted (Fig. 3c). At low membrane po-
tentials (<170 mV), the respiration was very low but
still may be said to be ohmic in that it was roughly
proportional to the membrane potential. The curves
for wildtype and mutator mitochondria are also near-
ly identical in this range, demonstrating that the in-
creased mutation rate has not influenced the proton
leak. However, at the higher membrane potentials, the
rate of oxidation was no longer under the control of
the membrane potential: only the (apparent) concen-
tration of substrate (i.e., succinate still not inhibited
by malonate) determined the rate of respiration. Such
steep curves forsuccinate oxidation versus membrane
potential have earlier been observed in other sys-
tems (liver mitochondria [51]) and similarly for glyc-
erol-3-phosphate oxidation in brown adipose tissue
mitochondria [52]. It is possible that our preparation
contained small amounts of mitochondrial membrane
fragments that may oxidize succinate unhampered by
the mitochondrial membrane potential; matrix-free
oxidation of succinate should be possible, as no soluble
electron carriers (NAD
+
) are needed for succinate (or
glycerol-3-phosphate) oxidation.
In the mtDNA mutator mitochondria, the appar-
ent breakpoint is decreased by as much as 32 mV, from
≈206 mV to ≈174 mV (Fig. 3c). This membrane potential
is apparently the highest that can be reached in these
mitochondria when the respiratory capacity is limited.
Thus, the lower level of membrane potential is not as-
sociated with high membrane proton leak but instead
reflects a diminished capacity for electron flux, and
Fig. 3.Proton leak analysis in wildtype and mtDNA mutator
mitochondria. Heart mitochondria were incubated principal-
ly as in Fig.1, with succinate (in the absence of rotenone) as
substrate, and oxygen consumption and membrane potential
were determined in parallel under identical conditions, after
addition of the indicated amounts of malonate. a)Oxygen con-
sumption. b)Membrane potential. c)Proton leak kinetics based
on the data in(a) and(b). The points are means ±SE from 7
independent mitochondrial preparations for each group.
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 287
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
the lowered membrane potential would explain why
ROS production is lower on reverse electron transport
in mtDNA mutator mitochondria.
The effect of rotenone.The inhibition of succinate
respiration observed above is thus due to endoge-
nously generated oxaloacetate. The accumulation of
oxaloacetate should be eliminated by the addition of
rotenone. Indeed, in the presence of rotenone, the res-
piration rate became sensitive to the membrane po-
tential, as the addition of FCCP now led to a marked
increase in respiratory rate (Fig. 2d) in the same mi-
tochondrial preparation. The succinate-supported re-
spiratory rate (state4) increased in both wildtype and
mtDNA mutator mitochondria in the presence of ro-
tenone (Fig. 2e). The addition of rotenone also almost
eliminated the difference in the levels of membrane
potential between wildtype and mtDNA mutator mito-
chondria (Fig. 2f) (the membrane potentials calculated
were now 207mV versus 199mV, i.e. the difference be-
came only 8mV).
Thus, we show here that mitochondria examined
under the conditions (succinate as substrate without
rotenone) widely used for studying the role of ROS in
longevity [42, 43] are in a bioenergetically unusual
state: the mitochondria are not in state 4, but a state
where respiratory capacity and membrane potential
are unphysiologically lowered, likely due to the ac-
cumulation of oxaloacetate in the mitochondria. The
membrane potential that could be maintained in both
types of mitochondria was thus unphysiologically low
with the mtDNA mutator mitochondria only able to
maintain an even lower potential (Fig.2b), which prob-
ably explains the lower ROS production in these mito-
chondria (Fig.1,a-c).
Obviously, under in vivo conditions, rotenone is
not present, and it may thus be considered strange to
claim that this condition, i.e., succinate plus rotenone,
is more physiological than just succinate. However,
to respire on only succinate is in itself unphysiologi-
cal, and when the NADH from malate oxidation can
be oxidized in Complex I, all oxidized succinate will
give rise to oxaloacetate, a product that inhibits suc-
cinate dehydrogenase. Under physiological conditions,
with fatty acids or glucose/pyruvate as substrate, the
oxaloacetate will condense with incoming acetyl-CoA
and thus not accumulate. Avoiding this accumulation
by using rotenone is thus an experimental condition
that– although it appears to be very artificial – more
closely mimics physiological conditions better than the
absence of rotenone.
Enhanced ROS production from mtDNA muta-
tor mitochondria energized by Complex I sub-
strates. Conditions such as those used above in the
classical design of mitochondrial ROS experiments, i.e.,
external succinate respiration and accompanying back-
ward electron flow, are rare in vivo [53-55]. We there-
fore continued by examining ROS production under
other, more physiologically relevant, conditions. Phys-
iologically, mitochondria are mainly energized with
ComplexI substrates and display forward electron flow.
To examine these conditions, we allowed mitochondria
isolated from heart or liver to respire on Complex I
substrates, pyruvate + malate (Fig.4a, shown for heart
mitochondria with Amplex Red system) or glutamate +
+ malate (Fig. 4b, shown for liver mitochondria with
DHE probe), and measured ROS production.
As seen (Fig. 4, a, b), the addition of substrate
led to an immediate and linear rate of ROS produc-
tion in both heart and liver mitochondria. When we
compared the rate of ROS production between mito-
chondria from wildtype and mtDNA mutator mice,
we found that, contrary to the case under classical re-
verse electron transport conditions (with succinate as
substrate), mtDNA mutator mitochondria from both
tissues investigated produced more ROS (heart 41%
and liver 70%) than wildtype mitochondria (Fig. 4, c, d).
Notably, oxygen consumption rates were not signifi-
cantly diminished in the mutator mitochondria un-
der these conditions [basal respiration (before FCCP)]
(Fig.4,e,f).
To identify the site of the increased ROS produc-
tion, we added rotenone to the mitochondria. Since ro-
tenone inhibits the flux of electrons from Complex I,
all ROS must then originate from Complex I (or from
upstream pyruvate or alpha-ketoglutarate dehydro-
genases). The addition of rotenone increased ROS
production in both wildtype and mtDNA mutator mito-
chondria (Fig.4, a-d), as expected [12, 40, 41]. After ro-
tenone addition, mtDNA mutator heart mitochondria
still had greater ROS production than wildtype heart
mitochondria, while liver mitochondria only showed
a non-significant trend toward an increase (Fig. 4, c
and d). This increased ROS production after rotenone
addition may indicate that upstream systems, e.g., py-
ruvate dehydrogenase [56], could be an important site
for ROS production in heart mitochondria, especially
in the mtDNA mutator mice.
It is clear from Figs. 1 and 4 that the magnitude
of ROS production observed with ComplexI substrates
compared to that observed with succinate is much
lower; Complex I substrate gave less than 20% of the
H
2
O
2
production obtained from succinate. That a very
high rate of ROS production is observed when succi-
nate is used as the mitochondrial oxidative substrate
probably explains why succinate is traditionally cho-
sen when examining ROS production ([9, 37, 57]). How-
ever, it is clear that qualitatively remarkably different
results were obtained when complexII and complexI
substrates were analyzed, yielding contrasting con-
clusions concerning the significance of mitochondri-
al mutations for ROS production. Thus, it would seem
that under conditions more resembling physiological
SHABALINA et al.288
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
Fig. 4.ROS production in mitochondria energized by ComplexI substrates. a,b)A representative trace of Amplex Red fluores-
cence(a) and dihydroethidium (DHE)(b) measured in wildtype heart(a) and liver(b) mitochondria energized by pyruvate+ ma-
late (a) or by glutamate+ malate(b). Additions were 0.2mg/ml heart (HM) or liver (LM) mitochondria, 5mM pyruvate for heart
mitochondria or 5mM glutamate for liver mitochondria (2mM malate was present in medium) and 1.7µM rotenone. c,d)Rate
of ROS production under conditions of forward electron flow supported by pyruvate+ malate in heart(c) and glutamate+ ma-
late in liver(d) mitochondria (measured principally as in a and b). The values represent means ±SE of 4-9 independent mito-
chondrial preparations for each genotype. *Statistically significant difference between wildtype and mtDNA mutator mitochon-
dria (p<0.05); #statistically significant effect of rotenone (p<0.05). e,f)ComplexI substrate-supported oxygen consumption
rates in heart and liver mitochondria from wildtype and mtDNA mutator mice. Values represent means ±SE of 4-9 independent
parallel mitochondrial preparations. *Indicates statistically significant difference between wildtype and mtDNA mutator mice
(p<0.05). ###Indicates significant effects of FCCP (p<0.001).
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 289
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
conditions, the mtDNA mutator mice do display a slight-
ly higher– not a lower– ROS production rate than do
wildtype mice.
Enhanced ROS production from mtDNA muta-
tor mitochondria energized by fatty acid-derived
substrates. In addition to pyruvate (and glutamate),
fatty acids are physiologically relevant substrates for
mitochondrial respiration, and the ROS production as-
sociated with fatty acid oxidation in the mtDNA muta-
tor mice is therefore of physiological relevance. Simi-
larly to what is the case for the classical Complex I
substrates (pyruvate, glutamate), fatty acid oxidation
occurs inside mitochondria and requires intact respi-
ratory chain activity and ATP production [58].
In contrast to other substrates such as malate,
glutamate, succinate or glycerol 3-phosphate, oxidation
of fatty acids requires no less than four enzymatic
reactions and donates electrons at several points in
the electron transport chain: Complex I, ETFQOR, and
ComplexII (via formation of succinate in the Krebs cy-
cle) [59]. Moreover, ROS production with fatty acid-de-
rived substrates is evident across a physiological range
of membrane potential and is relatively insensitive to
membrane potential changes [32,60]. This makes fatty
acid oxidation a good candidate for high rates of su-
peroxide or H
2
O
2
formation, due to possible leaks of
electrons to molecular oxygen at several different sites
independently of membrane potential.
Therefore, we examined how mitochondrial ROS
production with fatty acids as substrate is affected in
the mtDNA mutator mice. The fatty acid-derived sub-
strate palmitoyl-CoA, in the presence of carnitine, was
a suitable substrate for heart wildtype mitochondria,
providing a 10-fold increase in the rate of oxygen con-
sumption after FCCP addition, i.e., the maximal oxi-
dative capacity of heart mitochondria, exceeded the
basal respiration rate ten times (Fig.5a). Similar high
oxidative capacity was observed in wildtype heart mi-
tochondria supported by ComplexI substrates (Fig.4e),
indicating that, for the heart, fatty acids are equally
significant substrates as is glucose-derived pyruvate.
The rate of oxygen consumption in basal condi-
tions (not stimulated by FCCP) was not different be-
tween wildtype and mtDNA mutator mitochondria sup-
ported by palmitoyl-CoA + carnitine (Fig.5a), similarly to
Complex I substrates (Fig. 4e), and in contrast to suc-
cinate where mutant mitochondria had a lower rate
even in the basal state (Fig.2e).
Both mtDNA mutator mitochondria supported by
Complex I substrates or succinate had reduced spare
oxidative capacity (the FCCP response) as compared to
wildtype (Figs.4e and 2d), in agreement with [28]. Sur-
prisingly, high oxidative capacities supported by fatty
acid-derived substrates were equal in wildtype and
mutant mitochondria (Fig.5a). Thus, oxidation of fatty
acid substrates was not impaired, despite the severe
limitation in respiratory chain complexes in mtDNA
mutator mice, in agreement with [23]. Therefore, it
was interesting to see the high degree to which such
substrates affected ROS production.
The rate of ROS production was 29% higher in
mtDNA mutator heart mitochondria than in wildtype
mitochondria when fatty acid-derived substrates were
oxidized (Fig.5b), similarly to what we observed with
ComplexI substrates (Fig.4c) and in contrast to the re-
duced rate in the presence of succinate (Fig.1c).
For fatty-acid derived substrates, the addition
of rotenone increased ROS production in both wild-
type and mtDNA mutator mitochondria from heart
and the mtDNA mutator mitochondria had somewhat
Fig. 5.ROS production in mitochondria energized by fatty acid-
derived substrate (palmitoyl-CoA + carnitine). a) Basal and
maximal (after FCCP) oxygen consumption supported by pal-
mitoyl-CoA (30µM)+ carnitine (5mM) in heart mitochondria
from wildtype and mtDNA mutator mice. ###Indicates signifi-
cant effects of FCCP (p<0.001). b)Rate of ROS production sup-
ported by palmitoyl-CoA (30µM)+ carnitine (5mM) in heart
mitochondria from wildtype and mtDNA mutator mice, before
and after addition of rotenone. * Statistically significant dif-
ference between wildtype and mtDNA mutator mitochondria
(p<0.05). # Indicates significant effects of rotenone (p<0.05).
In (a) and (b), the values represent means ±SE of 5 indepen-
dent mitochondrial preparations for each genotype.
SHABALINA et al.290
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
higher ROS production than the wildtype mitochon-
dria (Fig.5b), similarly to what we observed with Com-
plexI substrates (Fig.4c).
Thus,in vitro, ROS measurement in mitochondria
supported by Complex I and fatty acid-derived sub-
strates yields results that may better reflect the situa-
tion in vivo [25]. In contrast, using succinate (without
rotenone) leads to mitochondrial inhibition by oxaloac-
etate and yields results contrasting thein vivoresults [25].
Enhanced ROS production in mtDNA mutator
mitochondria under conditions of ADP-stimulated
oxidative phosphorylation. Mitochondria in healthy
active cells would always be in some degree of state3
(active oxidative phosphorylation), although they may
not always be fully activated in this respect. To more
closely resemble the conditions in vivo, we therefore
studied mitochondrial ROS production (and respirato-
ry activity) in wildtype and mtDNA mutator mice mito-
chondria in the presence of high levels of ADP.
As expected, respiratory activity was increased
(Fig. 6a, thin blue line), whereas ROS production (Fig.6a,
heavy blue line) was reduced in wildtype heart mito-
chondria stimulated by ADP, in agreement with [61].
This ROS-attenuating effect of ADP was clearly less
pronounced in mtDNA mutator mitochondria (Fig. 6a,
heavy red line) than in wildtype mitochondria. Therate
of ROS production after the addition of ADP was thus
higher in the mtDNA mutator mouse mitochondria
than in wildtype mitochondria supported by palmi-
toyl-CoA+ carnitine (Fig.6a).
The rate of oxygen consumption in state 3 (after
ADP addition) was not different between mitochondria
from wildtype and mtDNA mutator mice when pal-
mitoyl-CoA was used as substrate (in agreement with
[23]) but with other substrates, the rate in the mutator
mitochondria was generally lower (in agreement with
[28]) (Fig.6b; Fig.S1 in the Online Resource1).
Similarly, to what is shown here for palmitoyl-CoA +
+ carnitine, the presence of ADP reduced ROS produc-
tion with other substrates– with the most pronounced
effect seen with succinate (without rotenone) (Fig.6c).
Notably, in the presence of ADP (state3), a higher rate
of ROS production in mutator versus wildtype mito-
chondria was evident in mitochondria energized not
only by palmitoyl-CoA + carnitine but also in mitochon-
dria energized by pyruvate + malate; the ROS produc-
tion rate was equal on succinate without rotenone
(Fig.6c; Fig.S1 in the Online Resource1).
To confirm these important observations obtained
in the Oroboros system, we also examined ROS produc-
tion in a plate-reader system. Also, here, mtDNA mu-
tator heart mitochondria showed higher rates of ROS
production with all three types of substrates tested
(Fig.6d). Remarkably, under these physiologically rele-
vant conditions of ADP-stimulated oxidative phosphor-
ylation, even succinate without rotenone tended to
show enhanced ROS production in mutant mitochon-
dria compared to wildtype mitochondria (Fig. 6c), in
clear contrast to the outcome without ADP and rote-
none (Fig.1). Thus, the very high ROS production rate
via reverse electron flux with succinate (without rote-
none) was drastically reduced in the presence of ADP
(Fig.6d).
Thus, under the most relevant physiological con-
ditions (active oxidative phosphorylation), mtDNA mu-
tator mitochondria produce more ROS than wildtype
mitochondria. The ROS production measured under
these conditions is thus in accordance with the results
of the measurementsinvivo[25].
Increased oxidative stress in liver from mtDNA
mutator in vivo. As mitochondria from mtDNA muta-
tor mice showed enhanced ROS production under phys-
iological conditions, we examined if this is associated
with an actual increase in oxidative stress in mtDNA
mutator mouse tissues. Therefore, we measured levels
of 4-hydroxynonenal adducts (HNE-adducts), a marker
of oxidative modification originating from lipid perox-
idation of proteins, in liver extracts (Fig.7,a-c). Wede-
tected both clear increases in the level of adducts of
specific molecular weights (specific proteins) (see e.g.,
band at ≈28 kDa in Fig. 7a) and a general increase in
the level of adduct formation (Fig. 7,b and c) in whole
liver lysate of mtDNA mutator mice (Fig. 7,a-c).
In addition to measuring 4-HNE-adducts, we also
examined the formation of protein carbonyls, gener-
ally used as a marker of oxidative stress [62]. Higher
levels of carbonyl groups were observed in lysates from
livers from mtDNA mutator mice as compared to wild-
type mice (Fig. 7,d ande), as expected both from the
4-HNE-adduct data above and from earlier studies [20].
We therefore conclude that at least in certain tis-
sues, an increased rate of mitochondrial mutations
may result in an increase in oxidative damage, proba-
bly mediated via an increased ROS production.
DISCUSSION
According to the mitochondrial hypothesis of ag-
ing [2, 3], aging is caused by mutations in mitochon-
drial DNA that would lead to alterations in the mito-
chondrially encoded proteins in the respiratory chain.
This would lead to increases in mitochondrial ROS
production. This would then constitute a vicious cycle
where the ROS thus produced would lead to further
mitochondrial mutations and thus further alterations
in the respiratory chain, leading to general oxidative
damage: aging.
However, clearly one prediction of the hypothesis
is that mitochondria with increased levels of muta-
tions in their DNA should produce enhanced amounts
of ROS. Until now this has not been demonstrated to
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 291
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
Fig. 6.ROS production in mitochondria stimulated by ADP. a)Representative traces of simultaneously measured oxygen con-
sumption (thin lines) and Amplex Red fluorescence (thick lines) in wildtype (blue lines) and mtDNA mutator (red lines) heart mi-
tochondria energized by palmitoyl-CoA + carnitine and then stimulated with ADP. Additions were 0.2 mg/ml mitochondria(Mit),
and 450 µM ADP. Simultaneous measurement of oxygen and fluorescence was done by O2k-Fluorescence Module. b) Oxygen
consumption rates in wildtype and mtDNA mutator mitochondria supported by palmitoyl-CoA + carnitine (Pal + Car) or pyru-
vate + malate (Pyr+ M) or succinate (Suc) as in(a); the effect of ADP addition. The values represent means ± SE of 5-7 indepen-
dent mitochondrial preparations for each genotype. * Statistically significant difference between wildtype and mtDNA mutator
mitochondria (p < 0.05). #, ### Indicates significant effects of ADP (p < 0.05 andp < 0.001). c) Rates of ROS production measured
simultaneously with oxygen consumption in wildtype and mtDNA mutator mitochondria supported by palmitoyl-CoA + carni-
tine (Pal + Car) or pyruvate + malate (Pyr+ M) or succinate (Suc) as in(a); the effect of ADP addition. The values represent means
±SE of 5-7 independent mitochondrial preparations for each genotype. *, ** Statistically significant difference between wild-
type and mtDNA mutator mitochondria (p < 0.05 andp < 0.01). ##, ### Indicates significant effects of ADP (p < 0.01 andp < 0.001).
d) Rates of ROS production under conditions of active oxidative phosphorylation. ROS production was measured using a
plate-reader simultaneously for all substrates including succinate in the presence of rotenone. 450µM ADP was present before
addition of mitochondria. The values represent means ±SE of 3 independent mitochondrial preparations. * Statistically signifi-
cant difference between wildtype and mtDNA mutator mitochondria (p < 0.05).
SHABALINA et al.292
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
Fig. 7. Oxidative damage in wildtype and mtDNA mutator mitochondria. a) Representative immunoblot analysis of 4-hy-
droxy-nonenal (4-HNE) adducts and corresponding Red Ponceau staining in liver whole tissue lysate from mtDNA mutator(M)
and wildtype mice (W) (15 µg protein/lane). b) Profile of 4-HNE-adducts in liver lysate. Lines indicate mean profiles of ad-
ducts examined in single or duplicate lanes as in figure(a) (heavyline for mtDNA mutator liver,n = 7;thinline for wildtype
liver,n = 5).Dashedline was obtained by subtraction of wildtype mean profile from mtDNA mutator mean profile. c) Relative
4-HNE-adducts amounts in muscle and liver lysate. Western blots as in(a) were quantified and the total area under the curve
was measured. The mean total amount of 4-HNE adducts in wildtype tissue was set to 100% and the levels of 4-HNE-adducts
in mtDNA mutator tissue is expressed relative to this. The values represent the means ± SE of 5-7 independent tissue prepa-
rations from wildtype and mtDNA mutator mice. d)Representative immunoblot of carbonylated proteins and corresponding
Red Ponceau staining of membrane in liver lysate. MSP,Mixture of Standard Proteins with attached DNP-residues. e)Relative
carbonylation in liver lysate. Western blots as in(d) were analyzed principally as in(c);n = 3. * Indicates statistically significant
difference between wildtype and mtDNA mutator mice (p < 0.05).
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 293
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
be the case, rather the opposite [20, 23, 24]. Due to the
importance of understanding the nature of aging, we
have here reexamined the effects of enhanced mito-
chondrial mutation rates on mitochondrial ROS pro-
duction. We find that the lowered rate of ROS produc-
tion earlier observed in the mtDNA mutator mice was
due to the traditional, succinate-based experimental
design used for the determination of ROS production.
We demonstrate here that when experimental condi-
tions are used that more closely approach physiolog-
ically relevant conditions, an enhanced level of ROS
production is indeed observed in the mitochondria of
the mtDNA mutator mice. Whereas these findings do
not prove that the manifestations of aging are caused
by increased ROS production due to increased mito-
chondrial mutations, they do reawaken this hypothesis
and thus open for further investigations into the enig-
ma of aging.
Observations of diminished ROS production in
mtDNA mutator mitochondria are caused by the
traditional experimental design. Earlier [23, 24], as
well as present observations (Fig.1) have consistently
found reduced ROS production in mtDNA mutator mi-
tochondria under the conditions traditionally utilized
for measuring mitochondrial ROS production. We have
here in detail analyzed the cause for this low ROS pro-
duction rate.
We found that when wildtype mitochondria were
oxidizing succinate under conditions when the oxida-
tion was not inhibited by oxaloacetate accumulation,
the mitochondria displayed ordinary respiratory con-
trol, i.e., respiration was stimulated by ADP and FCCP.
Under these conditions, a high membrane potential
was reached in the basal state (207 mV). As expected,
when mtDNA mutator mice mitochondria were exam-
ined under the same conditions, the respiratory capac-
ity was more than halved, leading to a slightly lower
membrane potential (199 mV). In these mitochondria,
the rate of ROS production in the presence of ADP was
actually slightly higher in the mtDNA mutator mice
mitochondria, in agreement with expectations that the
mutations in the mitochondrial genome resulted in
malfunctions in the respiratory chain components.
However, when the mitochondria were examined
as is traditionally done in ROS investigations, i.e., in
the absence of rotenone, the accumulation of oxalo-
acetate led to a strong inhibition of the succinate ox-
idation system. The respiratory rate was no longer
determined by the membrane potential, it is e.g., not
stimulated with FCCP, but it was limited by the low
(inhibited) activity of the succinate dehydrogenase.
Nonetheless, the respiration could raise the membrane
potential to 206 mV. This membrane potential is suffi-
cient to drive the electrons up to Complex  I, and a high
rate of ROS production occurred from this site. In the
mtDNA mutator mitochondria, the lowered capacity of
the respiratory system meant that it could only drive
the membrane potential up to 174 mV; this was insuffi-
cient to drive the electrons to Complex  I with the same
capacity. A lowered rate of ROS production was thus
observed, as reverse electron transfer is very sensitive
to the membrane potential (reviewed in [40, 41, 63],
also verified in brown-fat mitochondria [32]). The rea-
son for the lowered rate of ROS production seen in the
mtDNA mutator mitochondria is thus the artificial situ-
ation, with a profoundly inhibited activity of succinate
oxidation due to the accumulation of oxaloacetate, a
situation that does not occur normally in respiring mi-
tochondria; it may, however, occur in ischemia-reper-
fusion injury and succinate dehydrogenase deficiency
[53, 54]. The concept that mtDNA mutator mitochon-
dria display lower rates of ROS production in general
is thus an extrapolation from observations performed
under very special conditions.
Increased ROS production in mtDNA mutator
mitochondria with physiologically relevant sub-
strates and conditions. We find that, in contrast to
what we observe for succinate-derived ROS production,
ROS production derived from ComplexI substrates, fat-
ty acid substrates, and during active oxidative phospho-
rylation (with any substrate) was increased in mtDNA
mutator mitochondria, compared to wildtype mito-
chondria. Higher ROS production in isolated mitochon-
dria from mtDNA mutator mice is an important ob-
servation, implying that mitochondrial mutations can
indeed increase oxidative damage under physiological
conditions.
Divergent magnitudes of ROS production on Com-
plex  I- versus Complex  II-linked substrates are not
unique for the mtDNA mutator case described here.
In entirely different types of studies, a lower ROS pro-
duction from succinate but higher from Complex  I sub-
strates has been reported [64, 65]. In these studies, as
in ours, the lowered rates were associated with a re-
duced membrane potential [64,65].
ROS production supported by Complex  I sub-
strates in mtDNA mutator mice has earlier been in-
vestigated [19,  23, 24]. In contrast to our observations,
in these studies, ROS production was not higher in
mtDNA mutator mitochondria than in wildtype mito-
chondria. One reason for this discrepancy may be the
use of homovanillic acid for detecting ROS [19, 24].
Indeed, very high rates of hydrogen peroxide release
(≈  3000  pmol/min/mg on pyruvate) were reported,
values which are more than thirty times higher than
our values (≈  90 pmol/min/mg) and values reported
elsewhere (reviewed in [66]). Those observations may
therefore be ascribed to methodological differences.
However, Kukat etal. [23] found no difference between
wildtype and mtDNA mutator heart mitochondria de-
spite using very similar methods to the present paper,
but ROS production rate was also much higher in [23]
SHABALINA et al.294
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
compared to what is observed in the present article.
Wehave no explanation for these discrepancies.
Importantly we find that investigating ROS produc-
tion during ongoing oxidative phosphorylation is an es-
sential way to further approach physiological conditions.
Also during active oxidative phosphorylation,
the ROS production in mtDNA mutator mitochon-
dria is higher than in wildtype mitochondria.Mito-
chondria in normal cells are always at least to some
degree in state  3 (active oxidative phosphorylation),
and conditions where ROS production is studied in the
presence of ADP should thus better reflect the situa-
tionin vivo[13, 45]. We found that during active oxida-
tive phosphorylation, ROS production was higher in
mtDNA mutator mitochondria than in wildtype mito-
chondria, and this was the case with all different sub-
strates studied here (including succinate). This result
principally agrees with the observed increased ROS
production in mitochondrial state  3 (in the presence of
ADP) from normally aged animals as compared with
younger animals [13, 61].
A possible causative role of mitochondrial ROS
for cellular oxidative damage. The contribution of
ROS production to the aging process is a debated topic
[4, 67]. ROS can cause damage to proteins, lipids, and
DNA in the cell, but its role in the aging process is un-
clear. Mitochondria are the most important source of
ROS in the cell. It has been estimated that ≈0.4% of ox-
ygen consumed is converted to ROS during mitochon-
drial respiration [40]. However, the level of ROS is not
determined solely by the production rate but also by
the effects of e.g., antioxidants. Still, we find that the
lipid peroxidation rate in liver homogenates was en-
hanced in mtDNA mutator mice (Fig.  S2 in the Online
Resource  1), in parallel to the increase in mitochondri-
al ROS production generally described above in mtDNA
mutator mice. Accordingly, we found evidence for in-
creased oxidative damage in liver lysates from the
mtDNA mutator mice, in agreement with [20]. The ob-
served increases in 4-HNE adducts and protein carbonyl
groups are likely important indicators of the pathology
of mtDNA mutator mice. To what degree this reflects
what occurs during normal aging can of course not be
deduced from these observations in themselves, but
our results indicate that mutations in the mitochondria
may result in higher ROS production and higher ROS
levels that in their turn could induce further mitochon-
drial mutations and thus enhanced ROS production in
aself-reinforcing scheme.
CONCLUSIONS
This paper demonstrates that in contrast to what
has earlier been concluded from studies using a tradi-
tional experimental design for measuring ROS produc-
tion, the capacity for ROS production is increased in
mtDNA mutator mice when studies are performed un-
der conditions more closely approaching physiological
conditions. These enhanced rates of ROS production
in mtDNA mutator mice are associated with increased
signs of oxidative damage in tissues of mtDNA muta-
tor mice. The present studies are thus compatible with
hypotheses that suggest that increased levels of mi-
tochondrial mutations lead to enhanced rates of ROS
production and elevated oxidative damage of tissues,
and thus that aging processes may be understandable
as being the results of enhanced levels of mitochondri-
al DNA mutations.
Supplementary information.The online version
contains supplementary material available at https://
doi.org/10.1134/S0006297924020081.
Acknowledgments.The authors thank Aleksandra
Trifunovic for providing the initial breeding pair of
mtDNA mutator mice, Vladimir  Skulachev for stimu-
lating discussions, Anton  A. Tonshin for technical as-
sistance, and Sofie  Wagenius for establishing and veri-
fying mouse strains.
Contributions.BC, JN, and IGS conceived and de-
signed the work; experiments were performed by DE
and IGS (oxygen consumption and ROS), NG (oxidative
stress biomarkers), AVK and MYV (ROS), NP (immu-
noblotting), IGS (isolation of mitochondria and mem-
brane potential); MYV, NG, and IGS analyzed results;
DE, IGS, and JN wrote the manuscript. All authors re-
vised the manuscript and approved the final version.
Funding. This study was supported by grants
from the Swedish Research Council. AVK was support-
ed by a salary from the Academic Initiative of Stock-
holm University.
Ethics declarations. The experiments were ap-
proved by the Animal Ethics Committee of the North
Stockholm region.
Conflict of interest.The authors of this work de-
clare that they have no conflicts of interest.
REFERENCES
1. Sanchez-Contreras,M., and Kennedy, S.R. (2022) The com-
plicated nature of somatic mtDNA mutations in aging,
Front. Aging,2, 805126, doi:10.3389/fragi.2021.805126.
2. Harman, D. (2003) The free radical theory of ag-
ing, Antioxid. Redox Signal., 5, 557-561, doi: 10.1089/
152308603770310202.
3. Skulachev, V. P., and Lyamzaev, K. G. (2021) Mito-
chondrial Reactive Oxygen Species Aging Theory, in
Encyclopedia of Gerontology and Population Aging
(Gu, D., and Dupre, M. E., eds) Springer International
Publishing, Cham, 2021, pp. 3249-3256, doi: 10.1007/
978-3-030-22009-9_47.
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 295
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
4. Salmon, A. B., Richardson, A., and Perez, V. I. (2010)
Update on the oxidative stress theory of aging: does
oxidative stress play a role in aging or healthy ag-
ing?,Free Radic. Biol. Med.,48, 642-655, doi: 10.1016/
j.freeradbiomed.2009.12.015.
5. Lagouge,M., and Larsson, N.G. (2013) The role of mito-
chondrial DNA mutations and free radicals in disease
and ageing,J.Intern. Med.,273, 529-543, doi:10.1111/
joim.12055.
6. Brand, M. D. (2020) Riding the tiger – physiological
and pathological effects of superoxide and hydrogen
peroxide generated in the mitochondrial matrix,Crit.
Rev. Biochem. Mol. Biol., 55, 592-661, doi: 10.1080/
10409238.2020.1828258.
7. Edgar,D., and Trifunovic,A. (2009) The mtDNA muta-
tor mouse: Dissecting mitochondrial involvement in
aging,Aging (Albany NY),1, 1028-1032, doi:10.18632/
aging.100109.
8. Larsson, N. G. (2010) Somatic mitochondrial DNA
mutations in mammalian aging, Annu. Rev. Bio-
chem., 79, 683-706, doi: 10.1146/annurev-biochem-
060408-093701.
9. Capel, F., Rimbert, V., Lioger, D., Diot, A., Rousset, P.,
Mirand, P. P., Boirie, Y., Morio, B., and Mosoni, L.
(2005) Due to reverse electron transfer, mitochondri-
al H
2
O
2
release increases with age in human vastus
lateralis muscle although oxidative capacity is pre-
served,Mech. Ageing Dev., 126, 505-511, doi: 10.1016/
j.mad.2004.11.001.
10. Sohal, R.S., Arnold, L.A., and Sohal, B.H. (1990) Age-re-
lated changes in antioxidant enzymes and prooxidant
generation in tissues of the rat with special reference
to parameters in two insect species,Free Radic. Biol.
Med.,9, 495-500, doi:10.1016/0891-5849(90)90127-5.
11. Lass, A., Sohal, B. H., Weindruch, R., Forster, M. J.,
and Sohal, R. S. (1998) Caloric restriction prevents
age-associated accrual of oxidative damage to mouse
skeletal muscle mitochondria, Free Radic. Biol. Med.,
25, 1089-1097, doi:10.1016/S0891-5849(98)00144-0.
12. Mansouri,A., Muller, F.L., Liu,Y., Ng,R., Faulkner, J.,
Hamilton, M., Richardson, A., Huang, T. T., Epstein,
C.J., and Van Remmen,H. (2006) Alterations in mito-
chondrial function, hydrogen peroxide release and
oxidative damage in mouse hind-limb skeletal mus-
cle during aging, Mech. Ageing Dev., 127, 298-306,
doi:10.1016/j.mad.2005.11.004.
13. Chabi,B., Ljubicic,V., Menzies, K.J., Huang, J.H., Sal-
eem, A., and Hood, D. A. (2008) Mitochondrial func-
tion and apoptotic susceptibility in aging skeletal
muscle, Aging Cell, 7, 2-12, doi: 10.1111/j.1474-9726.
2007.00347.x.
14. Hansford, R. G., Hogue, B. A., and Mildaziene, V.
(1997) Dependence of H2O2 formation by rat heart
mitochondria on substrate availability and donor
age, J. Bioenerg. Biomembr., 29, 89-95, doi: 10.1023/
A:1022420007908.
15. Drew,B., Phaneuf,S., Dirks,A., Selman,C., Gredilla,R.,
Lezza, A., Barja, G., and Leeuwenburgh,C. (2003) Ef-
fects of aging and caloric restriction on mitochondri-
al energy production in gastrocnemius muscle and
heart,Am.J. Physiol. Regul. Integr. Comp. Physiol.,284,
R474-R480, doi:10.1152/ajpregu.00455.2002.
16. Simmons, R. A., Suponitsky-Kroyter, I., and Selak,
M. A. (2005) Progressive accumulation of mitochon-
drial DNA mutations and decline in mitochondrial
function lead to beta-cell failure, J. Biol. Chem., 280,
28785-28791, doi:10.1074/jbc.M505695200.
17. Papa, S., Sardanelli, A. M., Capitanio, N., and Picco-
li, C. (2009) Mitochondrial respiratory dysfunction
and mutations in mitochondrial DNA in PINK1 famil-
ial parkinsonism, J. Bioenerg. Biomemb., 41, 509-516,
doi:10.1007/s10863-009-9252-4.
18. Trifunovic, A., Wredenberg, A., Falkenberg, M.,
Spelbrink, J. N., Rovio, A. T., Bruder, C. E., Bohloo-
ly, Y. M., Gidlof,S., Oldfors, A., Wibom, R., Tornell,J.,
Jacobs, H. T., and Larsson, N. G. (2004) Premature
ageing in mice expressing defective mitochondrial
DNA polymerase, Nature, 429, 417-423, doi: 10.1038/
nature02517.
19. Kujoth, G. C., Hiona, A., Pugh, T. D., Someya, S., Pan-
zer, K., Wohlgemuth, S. E., Hofer, T., Seo, A. Y., Sulli-
van,R., Jobling, W.A., Morrow, J.D., Van Remmen,H.,
Sedivy, J. M., Yamasoba, T., Tanokura, M., Wein-
druch, R., Leeuwenburgh, C., and Prolla, T. A. (2005)
Mitochondrial DNA mutations, oxidative stress, and
apoptosis in mammalian aging,Science,309, 481-484,
doi:10.1126/science.1112125.
20. Trifunovic, A., Hansson, A., Wredenberg, A., Rovio,
A. T., Dufour, E., Khvorostov, I., Spelbrink, J. N., Wi-
bom,R., Jacobs, H.T., and Larsson, N.G. (2005) Somat-
ic mtDNA mutations cause aging phenotypes without
affecting reactive oxygen species production, Proc.
Natl. Acad. Sci. USA, 102, 17993-17998, doi: 10.1073/
pnas.0508886102.
21. Kolesar, J. E., Safdar, A., Abadi, A., MacNeil, L. G.,
Crane, J. D., Tarnopolsky, M. A., and Kaufman, B. A.
(2014) Defects in mitochondrial DNA replication
and oxidative damage in muscle of mtDNA mutator
mice,Free Radic. Biol. Med.,
75, 241-251, doi:10.1016/
j.freeradbiomed.2014.07.038.
22. Yu, T., Slone, J., Liu, W., Barnes, R., Opresko, P. L.,
Wark, L., Mai, S., Horvath, S., and Huang, T. (2022)
Premature aging is associated with higher levels of
8-oxoguanine and increased DNA damage in the Polg
mutator mouse, Aging Cell, 21, e13669, doi: 10.1111/
acel.13669.
23. Kukat, A., Dogan, S. A., Edgar, D., Mourier, A., Jaco-
by, C., Maiti, P., Mauer, J., Becker, C., Senft, K., Wi-
bom, R., Kudin, A. P., Hultenby, K., Flogel, U., Rosen-
kranz,S., Ricquier,D., Kunz, W.S., and Trifunovic,A.
(2014) Loss of UCP2 attenuates mitochondrial dys-
function without altering ROS production and uncou-
SHABALINA et al.296
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
pling activity,PLoS Genet.,10, e1004385, doi:10.1371/
journal.pgen.1004385.
24. Hiona, A., Sanz, A., Kujoth, G. C., Pamplona, R., Seo,
A. Y., Hofer,T., Someya,S., Miyakawa,T., Nakayama,C.,
Samhan-Arias, A. K., Servais, S., Barger, J. L., Porte-
ro-Otin,M., Tanokura,M., Prolla, T.A., and Leeuwen-
burgh,C. (2010) Mitochondrial DNA mutations induce
mitochondrial dysfunction, apoptosis and sarcope-
nia in skeletal muscle of mitochondrial DNA muta-
tor mice, PLoS One, 5, e11468, doi: 10.1371/journal.
pone.0011468.
25. Logan,A., Shabalina, I.G., Prime, T.A., Rogatti,S., Ka-
linovich, A. V., Hartley, R. C., Budd, R. C., Cannon, B.,
and Murphy, M. P. (2014) In vivo levels of mitochon-
drial hydrogen peroxide increase with age in mtDNA
mutator mice, Aging Cell, 13, 765-768, doi: 10.1111/
acel.12212.
26. Shabalina, I.G., Vyssokikh, M.Y., Gibanova,N., Csikasz,
R. I., Edgar, D., Hallden-Waldemarson, A., Rozhdest-
venskaya, Z., Bakeeva, L. E., Vays, V. B., Pustovidko,
A. V., Skulachev, M. V., Cannon, B., Skulachev, V. P.,
and Nedergaard, J. (2017) Improved health-span and
lifespan in mtDNA mutator mice treated with the
mitochondrially targeted antioxidant SkQ1, Aging,
9, 315-339, doi:10.18632/aging.101174.
27. Dai, D. F., Chen, T., Wanagat, J., Laflamme, M., Mar-
cinek, D.J., Emond, M.J., Ngo, C.P., Prolla, T.A., and
Rabinovitch, P. S. (2010) Age-dependent cardiomy-
opathy in mitochondrial mutator mice is attenuated
by overexpression of catalase targeted to mitochon-
dria, Aging Cell, 9, 536-544, doi: 10.1111/j.1474-9726.
2010.00581.x.
28. Edgar,D., Shabalina,I., Camara,Y., Wredenberg,A., Cal-
varuso, M.A., Nijtmans,L., Nedergaard,J., Cannon,B.,
Larsson, N. G., and Trifunovic, A. (2009) Random
point mutations with major effects on protein-cod-
ing genes are the driving force behind premature ag-
ing in mtDNA mutator mice,Cell Metab.,10, 131-138,
doi:10.1016/j.cmet.2009.06.010.
29. Silva, J. P., Shabalina, I. G., Dufour, E., Petrovic, N.,
Backlund, E.C., Hultenby,K., Wibom,R., Nedergaard,J.,
Cannon, B., and Larsson, N. G. (2005) SOD2 overex-
pression: enhanced mitochondrial tolerance but ab-
sence of effect on UCP activity,EMBOJ.,24, 4061-4070,
doi:10.1038/sj.emboj.7600866.
30. Udenfriend,S., Stein,S., Bohlen,P., Dairman,W., Leim-
gruber,W., and Weigele,M. (1972) Fluorescamine: a re-
agent for assay of amino acids, peptides, proteins, and
primary amines in the picomole range, Science, 178,
871-872, doi:10.1126/science.178.4063.871.
31. Kettle, A.J., Carr, A.C., and Winterbourn, C.C. (1994)
Assays using horseradish peroxidase and phenolic
substrates require superoxide dismutase for accu-
rate determination of hydrogen peroxide production
by neutrophils, Free Radic. Biol. Med., 17, 161-164,
doi:10.1016/0891-5849(94)90111-2.
32. Shabalina, I.G., Vrbacky,M., Pecinova,A., Kalinovich,
A. V., Drahota, Z., Houstek, J., Mracek,T., Cannon, B.,
and Nedergaard, J. (2014) ROS production in brown
adipose tissue mitochondria: The question of UCP1-
dependence,Biochim. Biophys. Acta, 1837, 2017-2030,
doi:10.1016/j.bbabio.2014.04.005.
33. Salvi,M., Battaglia,V., Brunati, A.M., La Rocca,N., Ti-
baldi,E., Pietrangeli,P., Marcocci,L., Mondovi,B., Ros-
si, C.A., and Toninello,A. (2007) Catalase takes part in
rat liver mitochondria oxidative stress defense,J.Biol.
Chem.,282, 24407-24415, doi:10.1074/jbc.M701589200.
34. Panov,A., Dikalov,S., Shalbuyeva,N., Hemendinger,R.,
Greenamyre, J.T., and Rosenfeld,J. (2007) Species- and
tissue-specific relationships between mitochondri-
al permeability transition and generation of ROS in
brain and liver mitochondria of rats and mice,Am.J.
Physiol. Cell Physiol., 292, C708-C718, doi: 10.1152/
ajpcell.00202.2006.
35. Shabalina, I.G., Hoeks,J., Kramarova, T.V., Schrauw-
en,P., Cannon, B., and Nedergaard,J. (2010) Cold tol-
erance of UCP1-ablated mice: Askeletal muscle mito-
chondria switch toward lipid oxidation with marked
UCP3 up-regulation not associated with increased bas-
al, fatty acid- or ROS-induced uncoupling or enhanced
GDP effects, Biochim. Biophys. Acta, 1797, 968-980,
doi:10.1016/j.bbabio.2010.02.033.
36. Nedergaard,J. (1983) The relationship between extra-
mitochondrial Ca
2+
concentration, respiratory rate,
and membrane potential in mitochondria from brown
adipose tissue of the rat,Eur.J. Biochem.,133, 185-191,
doi:10.1111/j.1432-1033.1983.tb07446.x.
37. Sahlin, K., Shabalina, I. G., Mattsson, C. M., Bakk-
man,L., Fernstrom,M., Rozhdestvenskaya,Z., Enqvist,
J. K., Nedergaard, J., Ekblom, B., and Tonkonogi, M.
(2010) Ultraendurance exercise increases the produc-
tion of reactive oxygen species in isolated mitochon-
dria from human skeletal muscle, J. Appl. Physiol.,
108, 780-787, doi:10.1152/japplphysiol.00966.2009.
38. Shabalina, I.G., Petrovic,N., Kramarova, T.V., Hoeks,J.,
Cannon, B., and Nedergaard, J. (2006) UCP1 and de-
fense against oxidative stress. 4-Hydroxy-2-nonenal
effects on brown fat mitochondria are uncoupling
protein 1-independent, J. Biol. Chem., 281, 13882-
13893, doi:10.1074/jbc.M601387200.
39. Skulachev, V.P. (1996) Role of uncoupled and non-cou-
pled oxidations in maintenance of safely low levels of
oxygen and its one-electron reductants, Q. Rev. Bio-
phys.,29, 169-202, doi:10.1017/S0033583500005795.
40. Murphy, M.P. (2009) How mitochondria produce reac-
tive oxygen species,Biochem.J.,417, 1-13, doi:10.1042/
BJ20081386.
41. Brand, M. D. (2010) The sites and topology of mito-
chondrial superoxide production, Exp. Gerontol., 45,
466-472, doi:10.1016/j.exger.2010.01.003.
42. Lambert, A. J., Boysen, H. M., Buckingham, J. A.,
Yang, T., Podlutsky, A., Austad, S. N., Kunz, T. H.,
ENHANCED ROS PRODUCTION IN MITOCHONDRIA 297
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
Buffenstein,R., and Brand, M.D. (2007) Low rates of
hydrogen peroxide production by isolated heart mi-
tochondria associate with long maximum lifespan
in vertebrate homeotherms, Aging Cell, 6, 607-618,
doi:10.1111/j.1474-9726.2007.00312.x.
43. Scialo,F., Sriram,A., Fernandez-Ayala,D., Gubina,N.,
Lohmus, M., Nelson, G., Logan, A., Cooper, H. M.,
Navas,P., Enriquez, J. A., Murphy, M. P., and Sanz,A.
(2016) Mitochondrial ROS produced via reverse elec-
tron transport extend animal lifespan, Cell Metab.,
23, 725-734, doi:10.1016/j.cmet.2016.03.009.
44. Brand, M.D., Goncalves, R.L.S., Orr, A.L., Vargas,L.,
Gerencser, A. A., Borch Jensen, M., Wang, Y. T.,
Melov,S., Turk, C. N., Matzen, J. T., Dardov, V.J., Pet-
rassi, H. M., Meeusen, S. L., Perevoshchikova, I. V.,
Jasper, H., Brookes, P. S., and Ainscow, E. K. (2016)
Suppressors of superoxide-H
2
O
2
production at site IQ
of mitochondrial complex I protect against stem cell
hyperplasia and ischemia-reperfusion injury, Cell
Metab.,24, 582-592, doi:10.1016/j.cmet.2016.08.012.
45. Goncalves, R. L. S., Quinlan, C. L., Perevoshchikova,
I.V., Hey-Mogensen,M., and Brand, M.D. (2015) Sites
of superoxide and hydrogen peroxide production by
muscle mitochondria assessed ex vivo under condi-
tions mimicking rest and exercise,J.Biol. Chem.,290,
209-227, doi:10.1074/jbc.M114.619072.
46. Fendel,U., Tocilescu, M.A., Kerscher,S., and Brandt,U.
(2008) Exploring the inhibitor binding pocket of re-
spiratory complex I, Biochim. Biophys. Acta, 1777,
660-665, doi:10.1016/j.bbabio.2008.04.033.
47. Boveris, A., and Chance,B. (1973) The mitochondrial
generation of hydrogen peroxide. General properties
and effect of hyperbaric oxygen,Biochem.J.,134, 707-
716, doi:10.1042/bj1340707.
48. Korshunov, S. S., Skulachev, V. P., and Starkov, A. A.
(1997) High protonic potential actuates a mechanism
of production of reactive oxygen species in mito-
chondria, FEBS Lett., 416, 15-18, doi: 10.1016/S0014-
5793(97)01159-9.
49. Fink, B. D., Bai, F., Yu, L., Sheldon, R. D., Sharma,A.,
Taylor, E. B., and Sivitz, W. I. (2018) Oxaloacetic acid
mediates ADP-dependent inhibition of mitochondri-
al complex II-driven respiration, J. Biol. Chem., 293,
19932-19941, doi:10.1074/jbc.RA118.005144.
50. Zeyelmaker, W. P., and Slater, E. C. (1967) The in-
hibition of succinate dehydrogenase by oxaloace-
tate,Biochim. Biophys. Acta,132, 210-212, doi:10.1016/
0005-2744(67)90214-8.
51. Nicholls, D. G. (1974) The influence of respiration
and ATP hydrolysis on the proton-electrochemical
gradient across the inner membrane of rat-liver mi-
tochondria as determined by ion distribution, Eur. J.
Biochem., 50, 305-315, doi: 10.1111/j.1432-1033.1974.
tb03899.x.
52. Nicholls, D.G. (1977) The effective proton conduction
of the inner membrane of mitochondria from brown
adipose tissue. Dependency on proton electrochem-
ical potential gradient, Eur. J. Biochem., 77, 349-356,
doi:10.1111/j.1432-1033.1977.tb11674.x.
53. Chouchani, E.T., Pell, V.R., Gaude,E., Aksentijevic,D.,
Sundier, S.Y., Robb, E. L., Logan,A., Nadtochiy, S.M.,
Ord, E.N., Smith, A.C., Eyassu,F., Shirley,R., Hu, C.H.,
Dare, A. J., James, A. M., Rogatti, S., Hartley, R. C.,
Eaton, S., Costa, A. S., Brookes, P. S., Davidson, S. M.,
Duchen, M. R., Saeb-Parsy, K., Shattock, M. J., Robin-
son, A. J., Work, L. M., Frezza, C., Krieg,T., and Mur-
phy, M.P. (2014) Ischaemic accumulation of succinate
controls reperfusion injury through mitochondrial
ROS,Nature,515, 431-435, doi:10.1038/nature13909.
54. Brockmann, K., Bjornstad, A., Dechent, P., Korenke,
C.G., Smeitink,J., Trijbels, J.M., Athanassopoulos, S.,
Villagran,R., Skjeldal, O.H., Wilichowski,E., Frahm,J.,
and Hanefeld,F. (2002) Succinate in dystrophic white
matter: a proton magnetic resonance spectrosco-
py finding characteristic for complex II deficiency,
Ann. Neurol.,52, 38-46, doi:10.1002/ana.10232.
55. Mills, E.L., Kelly,B., Logan,A., Costa, A.S., Varma,M.,
Bryant, C. E., Tourlomousis, P., Dabritz, J. H., Gottli-
eb, E., Latorre, I., Corr, S. C., McManus, G., Ryan, D.,
Jacobs, H. T., Szibor, M., Xavier, R. J., Braun, T., Frez-
za, C., Murphy, M. P., and O’Neill, L. A. (2016) Succi-
nate dehydrogenase supports metabolic repurpos-
ing of mitochondria to drive inflammatory macro-
phages, Cell, 167, 457-470.e413, doi: 10.1016/j.cell.
2016.08.064.
56. Starkov, A. A., Fiskum, G., Chinopoulos, C., Lorenzo,
B.J., Browne, S.E., Patel, M.S., and Beal, M.F. (2004) Mi-
tochondrial alpha-ketoglutarate dehydrogenase com-
plex generates reactive oxygen species, J. Neurosci.,
24, 7779-7788, doi:10.1523/JNEUROSCI.1899-04.2004.
57. Lambert, A. J., Buckingham, J. A., Boysen, H. M., and
Brand, M. D. (2010) Low complex I content explains
the low hydrogen peroxide production rate of heart
mitochondria from the long-lived pigeon, Columba
livia, Aging Cell, 9, 78-91, doi: 10.1111/j.1474-9726.
2009.00538.x.
58. Lesnefsky, E.J., Chen,Q., and Hoppel, C.L. (2016) Mi-
tochondrial metabolism in aging heart,Circ. Res.,118,
1593-1611, doi:10.1161/CIRCRESAHA.116.307505.
59. Perevoshchikova, I. V., Quinlan, C. L., Orr, A. L., Ger-
encser, A. A., and Brand, M. D. (2013) Sites of super-
oxide and hydrogen peroxide production during
fatty acid oxidation in rat skeletal muscle mitochon-
dria,Free Radic. Biol. Med.,61, 298-309, doi: 10.1016/
j.freeradbiomed.2013.04.006.
60. Seifert, E. L., Estey, C., Xuan, J. Y., and Harper, M.-E.
(2010) Electron transport chain-dependent and -in-
dependent mechanisms of mitochondrial H
2
O
2
emis-
sion during long-chain fatty acid oxidation, J. Biol.
Chem.,285, 5748-5758, doi:10.1074/jbc.M109.026203.
61. Lopez-Torres, M., Gredilla, R., Sanz, A., and Bar-
ja,G. (2002) Influence of aging and long-term caloric
SHABALINA et al.298
BIOCHEMISTRY (Moscow) Vol. 89 No. 2 2024
restriction on oxygen radical generation and oxi-
dative DNA damage in rat liver mitochondria, Free
Radic. Biol. Med.,32, 882-889, doi:10.1016/S0891-5849
(02)00773-6.
62. Dalle-Donne, I., Giustarini, D., Colombo, R., Rossi, R.,
and Milzani,A. (2003) Protein carbonylation in human
diseases, Trends Mol. Med., 9, 169-176, doi: 10.1016/
S1471-4914(03)00031-5.
63. Shabalina, I. G., and Nedergaard, J. (2011) Mitochon-
drial (‘mild’) uncoupling and ROS production: physi-
ologically relevant or not?, Biochem. Soc. Trans., 39,
1305-1309, doi:10.1042/BST0391305.
64. Bhattacharya,A., Lustgarten, M., Shi, Y., Liu,Y., Jang,
Y.C., Pulliam,D., Jernigan, A.L., and Van Remmen,H.
(2011) Increased mitochondrial matrix-directed super-
oxide production by fatty acid hydroperoxides in skel-
etal muscle mitochondria, Free Radic. Biol. Med., 50,
592-601, doi:10.1016/j.freeradbiomed.2010.12.014.
65. Iacono, L. L., Boczkowski, J., Zini, R., Salouage, I.,
Berdeaux, A., Motterlini, R., and Morin, D. (2011) A
carbon monoxide-releasing molecule (CORM-3) un-
couples mitochondrial respiration and modulates the
production of reactive oxygen species,Free Radic. Biol.
Med.,4, 4, doi:10.1016/j.freeradbiomed.2011.02.033.
66. Lambert, A. J., and Brand, M. D. (2009) Reactive oxy-
gen species production by mitochondria,Methods Mol.
Biol.,554, 165-181, doi:10.1007/978-1-59745-521-3_11.
67. Perez, V. I., Bokov, A., Van Remmen, H., Mele, J.,
Ran, Q., Ikeno, Y., and Richardson, A. (2009) Is the
oxidative stress theory of aging dead?, Biochim. Bio-
phys. Acta, 1790, 1005-1014, doi: 10.1016/j.bbagen.
2009.06.003.
Publisher’s Note. Pleiades Publishing remains
neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.